DeepSeek-V4-Pro
Specs & Pricing
API pricing per 1M tokens and core technical specificationsCost Calculator
Simulate estimated API usage cost based on your specific traffic and context sizeSimulation Results
Calculation Details
Alternative Models
Models with comparable benchmark score and pricing tier| Model | Provider | Context Window | Max Output | Training Cutoff | Input / 1M | Cache Price / 1M | Output / 1M | Benchmark | Speed | Total |
|---|---|---|---|---|---|---|---|---|---|---|
OpenAI👑 | 400K | 128K | 2025-08 | $0.75 | $0.075 | $4.50 | 87.5 | Fast | $0.0216 | |
xAI | 1M | - | - | $1.25 | $0.20 | $2.50 | 85 | Standard | $0.0217 | |
OpenAI | 200K | 100K | 2024-06 | $1.10 | $0.275 | $4.40 | 71.75 | Standard | $0.0268 | |
Anthropic👑 | 200K | 64K | 2025-07 | $1.00 | $0.10 | $5.00 | 69.5 | Fast | $0.0350 | |
OpenAI | 200K | 100K | 2023-10 | $1.10 | $0.55 | $4.40 | 62.75 | Standard | $0.0296 |
Provider
Benchmark
87.5Speed
Fast
Training Cutoff
2025-08
Context Window
400K
Max Output
128K
Input / 1M
$0.75
Cache Price / 1M
$0.075
Output / 1M
$4.50
Provider
Benchmark
85Speed
Standard
Training Cutoff
-
Context Window
1M
Max Output
-
Input / 1M
$1.25
Cache Price / 1M
$0.20
Output / 1M
$2.50
Provider
Benchmark
71.75Speed
Standard
Training Cutoff
2024-06
Context Window
200K
Max Output
100K
Input / 1M
$1.10
Cache Price / 1M
$0.275
Output / 1M
$4.40
Provider
Benchmark
69.5Speed
Fast
Training Cutoff
2025-07
Context Window
200K
Max Output
64K
Input / 1M
$1.00
Cache Price / 1M
$0.10
Output / 1M
$5.00
Provider
Benchmark
62.75Speed
Standard
Training Cutoff
2023-10
Context Window
200K
Max Output
100K
Input / 1M
$1.10
Cache Price / 1M
$0.55
Output / 1M
$4.40
All AI Models Pricing & Cost Simulator
Compare cutting-edge models from OpenAI, Anthropic, Google, and DeepSeek. Accurately simulate real-world API costs with prompt caching and custom context length.
DeepSeek-V4-Pro Frequently Asked Questions (FAQ)
Key insights into DeepSeek-V4-Pro API pricing, context capacity, caching savings, and performance benchmarks
Q. What is the API pricing for DeepSeek-V4-Pro? How much does it cost per 1M tokens?＋
DeepSeek-V4-Pro (DeepSeek) is priced at $1.3200 per 1 million input tokens and $3.9600 per 1 million output tokens. The output rate is approximately 3.0x higher than the input rate. You can simulate your exact monthly API expenses tailored to your request volume and token length using the calculator above.
Q. How does off-peak pricing work for DeepSeek-V4-Pro?＋
DeepSeek-V4-Pro offers discounted off-peak pricing during designated off-peak windows (平日 00:30-08:30 (UTC+8) & 土日終日). During off-peak hours, API rates are $0.6600/1M input (vs standard $1.3200) and $1.9800/1M output (vs standard $3.9600), with cached input at $0.0220/1M. Scheduling batch jobs and async processing during these hours yields significant cost savings.
Q. What are the context window and maximum output limits of DeepSeek-V4-Pro?＋
DeepSeek-V4-Pro features a 1M context window and a maximum output limit of 384K tokens per request. This massive 1M+ context capacity allows you to process extensive document libraries or entire codebases in a single prompt.
Q. Does DeepSeek-V4-Pro support Prompt Caching, and how much cost can be saved?＋
Yes, DeepSeek-V4-Pro supports prompt caching. Cached reads cost $0.0440 per 1M tokens, delivering an estimated 97% discount compared to standard input rates. (Minimum cache threshold: 64) Caching reusable system instructions and documentation substantially lowers recurring API expenses.
Q. What is the benchmark score (90) and performance profile of DeepSeek-V4-Pro?＋
DeepSeek-V4-Pro achieves an overall benchmark score of 90/100 with a "Standard" speed rating. It serves as a premier flagship model suited for complex reasoning, autonomous agents, and advanced coding.
Q. Which alternative models are comparable to DeepSeek-V4-Pro?＋
Key alternatives in similar price-to-performance tiers include GPT-5.4 mini, grok-4.3, o4-mini. We recommend comparing them based on your latency targets, context window needs, and prompt caching benefits using the full comparison tool on this site.